The second, and final (praise the Lord), round of Jakarta 2012 gubernatorial election will be held tomorrow. My eyes have seen too many chequered patterns recently, and I might as well add yet another post to the flame. I am more interested in playing the numbers rather than watching the debates, so let us begin the game of one million assumptions.

The baseline numbers for Foke and Jokowi, based on the first round results, are 34.05% and 42.60%, respectively. Let’s just assume that both camps retain these numbers as the baseline, although I suspect there will be more people to switch from the incumbent to the challenger than the other way around.

Next is those who voted for other contestants. For those who voted for Faisal Basri, they will never ever vote for the incumbent, and hence their choice will only be to spoil the vote or to vote for Jokowi. Let’s assume (you have been warned, there will be one million assumptions) that 20% of them are hardcore Faisal Basri supporters who will spoil the vote and not vote for Jokowi as they don’t want to be tainted with an indirect vote for Prabowo. And that, accounting for statistical error, another 5% somehow chose to vote for Foke, leaving Jokowi getting 75% share of Faisal Basri’s voters.

Next, Alex and HNW, the other two contestants whose parties endorse Foke for this round of election. Assuming that the party machinery is so strong and solid such that most of their voters follow the party line, Foke will get 90% share of these voters and Jokowi to get the rest.

For Hendardji, his 1.98% vote in the first round is basically a product of statistical randomness. It’s like basically anyone would be able to get that number if he/she was running in the election. So let’s just split his number equally between Foke and Jokowi.

Finally, let’s put the turnout to be slightly higher than the first round, which was at 63.62%, to be at 65%. And that Foke will get the same share of vote that he got in the first round if compared to Jokowi for these additional voters (i.e., 34.05%:42.60% = 44.42%).

With all these assumptions, Foke will get 50.41% while Jokowi will get 49.59%. A very close win for the incumbent that definitely will trigger vote recount and appeals.

All is well, then, for the incumbent? Not really. The bad news is the scenario that I was describing above is perhaps the *best* scenario for Foke. And if in the best scenario he could only win by 1 point, it means that he will be the underdog in this election. The numbers just don’t add up for him.

More realistically, you would expect that those that voted for Alex and HNW would not follow their party line so blindly. So even if we lower a little bit Foke’s share from those voters from 90% to 80%, Jokowi will already be winning by 2.5 points: 51.21% to 48.79%.

Another key for Jokowi is *turnout*. Basically he just needs to increase the turnout to win the election. The profile of these voters is more similar to Faisal Basri’s supporters than to the generic Jakarta population. They certainly don’t like the incumbent, but they just didn’t bother to vote. The more people vote, the higher chance Jokowi will win.

(Jakarta, though, has a what I call as structural spoilt vote problem, perhaps around 10-15%, due to the following reasons: high number of Jakartans living overseas, travelling businessmen, and those who still retain their Jakarta’s IDs although they are already living in Bodetabek suburbs like Serpong or Karawaci.)

The following will be my wild guess for tomorrow (more realistic, no?). Only 10% of Faisal Basri’s supporters will spoil the vote. Jokowi will get 30% share of Alex’s voters, but only 20% from HNW’s due to stronger PKS machinery. Turnout will increase by 5%, and Foke will only get 30% share out of these additional voters. And there will be more Foke’s voters who will switch to Jokowi rather than Jokowi’s to Foke. The final result turns out to be a 8-point win for Jokowi.

(For a 60-40 win, Jokowi needs to get half of Alex-Hendardji-HNW voters and 20% of Foke voters, which is less likely.)

A caveat, though. All of these, of course, come under the assumption that the election will be clean.

(After I posted this, I just realized that it should be “Spoilt vote” rather than “Spoil vote”, but I just didn’t bother to fix it. And I realized that the images are bad, so you might want to click them to see them clearly.)